What happens if stainless/carbon welding was done with E70xx

Date 10-07-2010 16:43

"Hypothetical" question:

In what manner would a weld fail if I had an A36 doubler plate welded to a Stainless steel tank using E70 type filler metal?
(I probably triggered a lot of alarm bells so far)
Say that the doubler is to be used to reinforce a tank for a padeye and that the pick has already been done with no failure.
Exact grade of stainless is unknown - scanned drawings state SA240; the type is unknown but is likely 304 or 316. The weld is a 5/16" fillet all around - all I know is that it was carbon steel wire - I don't know the process or exact fillet metal.

This is not our mistake. I was asked for an opinion, which is: Even though one lift was made succesully, the weld could still fail so it should be removed, tank checked for soundness, and replace the weld with 309L filler metal. However, I don't have anything that says why the weld is not correct except that all recommendations are to use 309L filler metal.

Some folks have been able to point to similar situations in which the welds did not fail, but were replaced at the next opportunity. So that is why I'm wondering about this. Is it automatically a "failure waiting to happen" or is it simply "not recommended but might be OK"?
Thanks.

By ravi theCobra (**)

Date 10-07-2010 17:03

I would say the most important consideration is SERVICE.

If this is just a lifting eye it might pick up the load 5 or 6 times in its entire service life and you can mark it MPT and
LPT before use -

However if it is going to be used for moving the tank once a month or week that brings up another " NONHYPOTHETICAL "
question -

Lots of stainless steel ( P - 8 ) is welded with 7018 and serves its service life and no one ever thinks about it -

By jwright650 (*****)

Date 10-07-2010 17:21

I welded a SS lifting handle back on my outdoor grill with E71T-11, and it's held up all of these years....LOL, OK, I'll go back to my corner and be quiet.
Sorry Chet, I'm no help.

By eekpod

Date 10-07-2010 18:47

I love these hypothetical questions.. they sound so real :)

By 803056

Date 10-08-2010 03:18

Hello Chet;

The problem that comes to mind is whether there is too much or too little ferrite in the weld to prevent cracking. I suspect if there was going to be a problem it would manifest itself during the welding of the carbon to stainless. Too little ferrite would result in cracking problems when those elements with limited solubility in the austenitic grains are rejected resulting in fissures that could lead to failure when subjected to sustained loads. The loads could be residual stresses or applied design loads.

Excessive ferrite could result in Sigma phase, which could embrittle the HAZ of the stainless. Small amounts of Sigma are all that is needed to embrittle the HAZ.

You might consider testing the weld and adjacent HAZ with fluorescent dye penetrant to see if you can detect any cracks that might be present. Even if no cracks are detected, it is no guarantee there is not a problem, but it would provide some level of assurance the welds are sound.

You might consider using Figure 1.1 found in D1.6-2007, Annex I to approximate the microstructure of the weld and HAZ. That should provide a good approximation of the expected ferrite and whether martensitic structures can be expected in the HAZ. A better approximation can be made using the Kotecki/Lippold diagram.

Best regards - Al

By ssbn727 (*****)

Date 10-08-2010 06:58

As usual, excellent advice there Al. ;)

Respectfully,
Henry

By js55

Date 10-08-2010 10:32

Chet,
The problem is martensite. When using carbon steel filler metal in a dissmilar between carbon steel and stainless steel the filler metal will be diluted by the SS and the result will be a low allow weld deposit that when cooled can form martensite and therefore be embrittled. Unheat treated it can be hard as a rock, with no toughness at all, and as crack sensitive as any martensitic deposit. The reason for using 309 is that its alloy content is bumped up (from 308) in order to accomodate this dilution. You can even get martensite with 308 if you have high dilution (for example in a GTAW root pass) thats why 309 is recommended. CS filler is far worse than 308.

By 803056

Date 10-08-2010 17:02

JS55 is on the money with regards to the formation of martensite in the partially melted zone at the weld interface.

Think of the carbon equivlency formulas. The elements considered by the CE formula includes Cr, Ni, Mn, Si, etc. Each contribute to the CE. The higher the CE, the more likely martensite will form. The truth of the matter is, there are several CE formulas, each provides a little different CE, but the results is the same, high CE indicates the weld or partially melted zone in the HAZ (against the molten weld puddle) will likely form martensite when cooled quickly. Considering the high percentage of Cr in the austenitic stainless, alloyed with the constiuents of the carbon steel filler metal, almost ensures martensite is going to be a problem.

Best regards - Al

By Shane Feder (****)

Date 10-09-2010 04:17

Hi guys,
Just to support what Jeff and Al have stated with a practical example.
Got called out a while back to investigate major cracking (360 degrees) on a 2" coupling that had been welded onto the side of a hydraulic oil tank.
It was a huge Thyssen Krupp Stacker / Reclaimer that was being constructed at a Coal Loading Facility.
The tank was 3 metres square and was made of 6 mm 316 stainless steel. For some reason it had been painted blue so when the welder was requested to add a 2" coupling to the side of the tank he just presumed it was mild steel. Cut the hole (with a hole saw) and welded the coupling in with E7018 electrodes.
Cracking occured at the weld toe for the whole circumference of the weld.
A good learning experience for me (and definitely the welder)
Regards,
Shane

By Milton Gravitt (***)

Date 10-10-2010 14:49

That's why I keep a magnet around if I have any question,yes I know some SS is magnetize.

M.G.

By CHGuilford (****)

Date 10-11-2010 16:25

Thank you everyone for the feed back.

At this point, I'm not sure what will become of this "hypothetical" situation. But the encouraging part is that they asked the right questions of many folks, so I think they will address te problem properly.

It's a mill shutdown situation and many planned activities aren't going according to plan. I think the one I described is relatively minor compared to finding that existing conditions uncovered during demo are much worse than was anticipated.